HAL Errors of Conclusion or Functioning?

"It seems to me that this criticism depends on a confusion between two kinds of mistake. We may call them 'errors of functioning' and 'errors of conclusion'. Errors of functioning are due to some mechanical or electrical fault which causes the machine to behave otherwise that it was designed to do... Errors of conclusion can only arise when some meaning is attached to the output signals from the machine."

(449 Turing)

This passage brings me back to the scene in 2001 Space Odyssey when HAL remarks that "human error" is the only explanation for the discrepancy between the two HAL 9000's different results. At first, I thought this was just a defensive remark made on HAL's part - a prideful reluctance to admit he had made any sort of mistake. However, his statement about human error could also be read as him acknowledging an error of functioning within himself as a result of some mechanical/electrical fault caused by the humans who created him. Under this circumstance, HAL is not malfunctioning at all but is rather still operating without errors according to his system - perfectly performing the tasks he is capable of doing according to his programming. Rather, I don't know if it's possible for him to recognize it as an error in himself at all - HAL may only recognize that in the event of a discrepancy between two HAL units, the fault lies with the human element in the equation - HAL's programmers and technicians.
For some reason, I find it easier to lay the fault at the feet of the programmers than the machine - perhaps that means I'm the type who would blame the parents for the outcome of the child.
The other explanation for HAL's behavior is that an error of conclusion occurred. Honestly, I'm still a little shaky on the exact meaning of an "error of conclusion, and Turings explanation moves a little too quickly for me to really grasp it. I'd love some help interpreting it so that I can further attempt to understand the potential reasons behind HAL's errors (if in fact they are errors at all).
-James G.

2001: A Space Odyssey

It is interesting to note Hal's transition, albeit quick, from trustworthy friend to murderous foe. Look, for instance, at the way he is first introduced and talked about by the Dave and Frank. He is a super intelligent machine, he never makes mistakes, he is integral to the success of the mission, he is just another crew member. It seems that the partnership between the humans and this machine is rock solid, they allow Hal to be part of the crew and Hal does whatever they tell him to. Perfect, right?

But when you think about it, Hal doesn't actually go through a transition at all. He doesn't change at the hands of some troublesome event. The same computer that aided the mission and kept the slumbering astronauts alive is the same one that eventually killed them. The only thing that changed was the level of threat felt by his human counterparts.

It is easy to befriend someone, or something, that doesn't inherently threaten you. It is even easier when this being is programmed to be subservient to you. When asked if Hal could actually think, a political answer was given - "who can really say?". But we can tell that whether or not he can think is irrelevant to them. They just want him to do his job and assist them. They don't care if he's thinking or not. I think by the end of the film we can surmise that Hal probably can think.

This idea also jumped out at me from Turing's piece. He doesn't talk about it in much depth but he hints at this concept of threat. When he talks about the interrogator game, he mentions that a person would never be expected to win a race against an airplane, or an arithmetic challenge against a supercomputer. Certainly we agree, these machines are made to do these things at a much faster speed than humans ever could. But the other side of this coin is that humans expect that we should not be challenged by machines in terms of thinking and, well, acting human.

Perhaps there is some level of threat inherent in the idea of a computer being able to think or act like a computer. Maybe it's because of all the science fiction movies out there, like 2001, that show what could potentially happen when we bestow upon machines our heightened levels of consciousness. Since we have already lost the battles of productivity to machines, how would we feel to know we also lost the creative one?

Yes, but are they really in there??

I just don’t see how it could possibly come back to anything but a question of consciousness. From my perspective in 2013, it seems inevitable that a computer will eventually be able to trick his interviewer with some incredible algorithm. Even now, Siri may not be the brightest, but she definitely has a personality. (And it was no trouble at all for me to identify her as “she.”) Is this really the tricky problem Turing and others imagined? He addresses consciousness as one of the possible forms of objection, but only so far as to say that some people argue that consciousness is required to carry out a sufficiently convincing conversation for their interviewer. While I don’t share that view, I believe that “thinking” means something very different from “processing.” See: Chinese Room. Turing changes the “can machines think” question into a test that seems far less relevant and interesting. I don’t know the cultural context that this paper was published in, and I hardly want to accuse Alan Turing of missing the point, but his question seems far less deep and meaningful that “Can machines think?”

This article reminded me of a Star Trek episode where they ended by deciding that an android functioned as a member of the crew, not as a piece of equipment. The android, Data, regularly passed the Turing test and even had to convince people that he was not, in fact, human. He had a thinking face and made insightful and self-reflective statements, but at his core was simply a processor. It is strange to think that we are, in a sense, biological processors at our cores, but somehow we do find ourselves to be conscious creatures. I just can’t accept that any digital computer could possibly feel the same awareness I do. I didn't intend for this post to get so personal, but I suppose that's in the nature of the question.

Turing influencing Kubrick

Almost every time somebody writes a novel or makes a movie about computers or robots in the future, artificial intelligence is assumed. HAL 9000 in 2001: A Space Odyssey is a global icon for the potential terrors of artificially intelligent computers.  AI is often associated with faulty and a sometimes blatantly evil mentality. This probably reflects an understandable discomfort with the idea of sharing a similar intelligence with something that isn’t human (even if it is created by humans). There’s the strong theme in many fictional works that the AI means learning, and when it learns, it may become smarter than humans,  know more than humans, and is therefore dangerous.

It is hard to say that Kubrick's idea was not inspired by the work of Alan Turing. About 20 years before the film was released, Turing brought up the controversial question "Can machines think?" We remember Turing from earlier in the class through his Turing Machine. Though the his machine was never successfully produced or marketed, it was important in the history of technology because of its unique method. As Edwards states, "Turing proved that this extremely simple machine, given the appropriate finite "configurations" or instructions could solve almost any...symbolic problem that had a solution (242).


Years later when Turing posed his question about thinking machines, he understood that he would need  empirical evidence for this and proposed what is known as the Turing Test which hoped to determine if a machine was capable of thinking. The test studies whether a blind interrogator can determine the difference between a response from a computer or a human. The idea was that if the questioner could not tell the difference between human and machine, the computer would be considered to be thinking. Turing addresses many issues that have been brought up with his idea and tries to find fallacies in them, but at the end of his essay admits that he has "no very convincing arguments" to support his views. That being said, his question has started a conversation that we are still having today and that most likely influenced Kubrick's film and its groundbreaking success.

The Leap to Knowledge: Deep Reading

Wolf and Barzillai's article on deep reading showed me that online reading can lead to deep reading: by learning to read selectively.  I have been reluctant to make the shift from reading from a book to reading a book online for fear that the quality of knowledge might be compromised.  I can spend the same amount of time reading online as I do with a book, by making the right choices and distinctions from what is distracting and what is real reading for me.

This article helped me define the meaning of what it is to acquire knowledge versus mass information.  The Aristotle analogy of the three lives is a perfect reference to achieving knowledge:

"Of the three lives Aristotle speaks of, the life of action, the life of enjoyment, and the life of      contemplation, we have the two, but we lack the other, contemplation. That, I thought, is why ours is a violent city." [First Page]

Deep reading is critical to our development of knowledge. There is a possible misconception in cyberspace that the more information scanned through and amassed, the more you know, but what about the quality of that "knowledge"?  Take for instance spellcheck and grammar check in Microsoft Word, the rules are still being taught in grammar school, but with all these nice electronic tools, why go through the trouble of learning to spell if you can always rely on spellcheck?  Did the student/writer really learn?  Ah! That's where the concept of deep reading comes in.

Deep reading, like correct spelling and grammar, are practices that require contemplation to help make the connections and construction of knowledge; the smaller child needs more time than an adult to learn to construct knowledge, but once this is achieved, deep knowledge starts to set in, is my perception of the author's article. The processes they refer to that make up deep reading require time, at least for me.  Deep reading processes help produce comprehension, reflection, induction, deduction, analogical skills, critical analysis, and insight, and ultimately, help produce an expert reader. These processes cannot be achieved by competing with time;  it takes time to absorb things and then synthesize them.

I believe we need to learn how to think and discern in order to achieve good judgment and make sound decisions.  This is the challenge of the contemporary educator, to help foment all of these processes as early as possible so that students can achieve deep reading to perform at a higher level, and it can be achieved using both traditional print and electronic print.

AI, One Day in Our Reach

When I think of AI as a companion or some sort of life form which is synthetically like a human, the first thing that comes to mind is Commander Data from the ageless television series Star Trek: The Next Generation. Commander Data, an AI android, has no emotion and has incredible processing speed to calculate and solve problems. However, his goal his to become more human because he lacks experiences humans develop over their lives. The point here is that Data is what people who in Edwards paper wanted to develop, a synthetic brain that somewhat mimics what the human brain. The hardware brain may be possible and perhaps turn out like Data, but one thing the show provided was that the hardware brain may not understand implicit human interactions. Guessing a behavior is not a mathematical equation, but an intuition that we have experienced in the past ourselves. The units in the hardware can gather information by making connections such as a neuron. One of the problems is that we have the spinal cord, which acts as a second brain for our motor skills, can be difficult for the AI brain to recreate.

            As of today, Edward’s paper is true about how AI one day can take command and control themselves without human intervention. For example, Watson that IBM created which can answer any question given to it. However, the mathematical program/code is too discrete for the mind to be mimicked. Determinism is a concept that many people believe that what ever we think, feel, and do is already determined by the course of actions in that particular situation. This concept can be easily be applied the idea that our human brain can be copied into software. However, on the other spectrum, some people believe in the free will, that nothing is determined or preset. This may be harder for an AI to mimic and perhaps too complicated for a AI to even understand the magnitude of coding that needs to go into it. To make an AI that is a tool for human enhancement is possible today, but to have an AI self aware is perhaps still far from our reach.

Imperfect Intelligence

Reading through Edwards was at times, dense and a little bit too detail obsessed. It wasn't until the conclusion that it really hit me- how can we as a society, truly believe that we can create a sort of artificial intelligence that exceeds our own? We've all seen those doomsday scenario films and Sci-fi shows that provide an eerie prophecy of a world where computers have taken over the world. These films have raised legitimate concerns regarding the implications of giving too much power to machines and depending on them to the point where they are constant legislators in our lives. Considering just how much we know about the human brain, it has become understood that creating artificial intelligence is tangible. I don't disagree, but I think what is often overlooked is just how much we may not know about the human brain. Edwards suggests at the end of his article that "artificial intelligence man-computer symbiosis, and human information processing represent the reductions necessary to integrate humans fully into command and control" (Edwards 273). What this meant to me, was that the different technological innovations made and the stride towards artificial intelligence are essential to integrate humans entirely into commanding and controlling these operations. Moreover, going back to the fallible concept, that is the human brain, when we make these machines, we expect them to have a certain kind of perfection that the human mind can't achieve. This standard of perfection is an illusion in a world where we don't have a perfect understanding of the brain or of intelligence.

There is something to be said about the way we view computers and artificial intelligence. We expect so much of them and we have a hard time accepting when they fail us. We set machinery to a higher standard than we hold ourselves. I'll be the first to say that when Macbook refuses to turn on or fails to recognize the wifi networks in the area, I feel a much greater sense of disappointment than when my friend forgets we have a Skype date. If you ask me, that's problematic. We are unable to handle when technology fails us because we have unrealistic expectations of them. What society too often forgets is that machines aren't perfect and that there is no such thing as an intelligence superior to our own because we hold the reins to the amount of power given to machines.